Makai P.

GitHub        LinkedIn

Located in Florida

Interested in mechanistic interpretability, non-euclidean geometry, and mid-century graphic design. I volunteer doing neuroscience research and oceanographic research, but in my free time I like to make art, music, and code websites like this one.

Max Planck

June 2025—June 2026

At the Max Planck Florida Institute For Neuroscience, I worked in the Tian Lab, using generalized linear models (GLMs), long-short term memory (LSTM), and convolutional neural networks (CNNs) to quantify relationships between fiber-photometry-recorded serotonin (5-HT) measurements and fear responses in the ventral tegmental area (VTA) and orbitofrontal cortex (OFC). I also developed tools to automate behavioral analysis and reproject animal movement.

Harbor Branch

June 2024—Present

In the Geochemical Remote Sensing Lab at the Harbor Branch Oceanographic Institute, I develop machine learning approaches to quantify sediment-derrived dissolved organic carbon (DOC) in the Northern Gulf of Mexico (NGOM) using remote sensing (RS). We use colored dissolved organic matter (CDOM) to locate DOC, and use features such as Chlorophyll-a (Chl-a) and sea surface temperature (SST) to distinguish its source.

URIKA [bioworks]

June 2024—November 2024

I invented AWS Neptune's first data processing pipeline for NLP-derived biomedical relational databases using Python, Bash, and SPARQL. This was contrary to the conventional use of Neo4j and Google Cloud, and I drafted the AWS whitepaper for the project.